In the recent Viewpoint by Ebach and Holdrege, “More Taxonomy, Not DNA Barcoding” (BioScience 55: 822–823), the authors doubt that the application of DNA barcoding (the use of short nucleotide sequences as molecular markers for species identification) will really become a serious alternative to more conventional taxonomic techniques. They claim that science needs more taxonomists, not more barcoders. They take into consideration that “barcodes cannot reveal the types of their corresponding real species.”
I am afraid that this criticism represents more a misconception of what a “real species” is than a concern about the quality of forthcoming taxonomic practices. Diagnosing a species should not be confused with the ontological “thing” that a species really is in nature. Evolution is an ongoing change, and so is speciation. A completed species does not exist. Knowledge of the “real species” is one goal; operational diagnosis is another. Merging both goals into a single concept brings to light an obvious immanent conflict between these two desires. Whenever the criteria for species classification and identification are optimized, the ontological status of real species will become fuzzy, and vice versa.
The emphasis on diagnosis forces taxonomists to treat species as classes, not as things or individuals that exist as real entities in nature. Classes, however, have a very different ontological status than realities. The generation of classes depends on our current technology, and our skills for identifying similarities and differences will change with time.
One thing is clear: Barcoding is a method for recognizing differences among individual organisms. In this respect, it is no different from any other technique used in taxonomy, for example, counting the number of bristles on the legs of certain beetles. The call for “more taxonomy, not DNA barcoding” is comparable with a call for “more taxonomy, not bristle counting.” However, not all characters are suited for dividing organisms into biologically meaningful groups. Preexisting ancestral intraspecific polymorphisms have to be recognized. Does the barcode pattern covary with other taxonomically relevant characters? This question, among others, has to be scrutinized if barcoding is put on the test bench. If barcoding passes these hurdles, then taxonomy needs more bar-coders, and barcoding should compete with conventional methods for funding, since its pragmatic advantage is evident.
